A method and model for predicting the net energy value of feed raw materials for growing pigs and its application

Abstract

The present invention provides a method, model and application for predicting the net energy value of feed raw materials on growing pigs, the method comprising: 1) obtaining conventional chemical components and net energy values ​​of multiple feed raw materials; 2) clustering the feed raw materials according to the conventional chemical components of the feed raw materials; 3) performing supervised classification according to the clustering results to obtain a classification standard; 4) classifying the feed raw materials according to the classification standard, and using the conventional chemical components of the feed raw materials as input variables, and using the net energy value of the feed raw materials on growing pigs as output variables, to establish a prediction model for at least one category. The present invention classifies feed raw materials based on conventional chemical components and can well predict their net energy values ​​on growing pigs, which helps to solve the difficulties of expensive equipment, high professional requirements, large workload, and long time consumption required for determining the net energy value of feed raw materials, and further promotes the realization of accurate formulation of growing pig feed.

Description

Technical Field

[0001] The invention relates to the fields of animal nutrition and feed science, and in particular to a method and a model for predicting the net energy value of feed raw materials on growing pigs and their application. Background Art

[0002] The index system for measuring the effective energy value of pig feed raw materials mainly includes digestible energy, metabolizable energy and net energy system. Compared with digestible energy and metabolic energy systems, the net energy system can more truly reflect the energy utilization efficiency of feed, and the net energy system is the only effective energy system that considers the energy requirements of animals and the energy value of feed on the same basis. However, the determination of net energy is more complicated than digestible energy and metabolizable energy. It is generally determined by indirect calorimetry or comparative slaughter method on pigs. The former requires the use of a special respiratory calorimetry chamber, and the latter requires professional slaughtering and cutting equipment and personnel. Both methods are time-consuming and labor-intensive, which limits the development of the net energy value determination of feed raw materials and is not conducive to the promotion and application of the net energy system.

[0003] At present, although there are a few studies on the prediction model of the net energy value of a single feed raw material in growing pigs, it is only applicable to a specific type of feed raw material, with a small application range and low practicality. There are many types of feed raw materials commonly used in production, and a single prediction model is not sufficient to accurately predict the net energy value of all types of feed raw materials. Therefore, the feed raw materials can be divided into different categories based on the similar properties of conventional chemical components among feed raw materials, and a prediction model of the net energy value of feed raw materials with a wider range of applications can be established based on the raw material categories. In traditional feed science, the main feed raw materials are artificially divided into protein feed raw materials, energy feed raw materials, and fiber feed raw materials, etc. However, the classification indicators based on which each category is divided are relatively simple, resulting in great differences in the nutritional characteristics of feed raw materials in the same category, and the classification results of some feed raw materials are not accurate.

[0004] Therefore, there is an urgent need to develop a method that can quickly and accurately classify feed raw materials and then predict their net energy value for growing pigs. More specifically, it is necessary to establish a method to objectively classify feed raw materials based on their conventional chemical components, and further use conventional chemical components as model parameters to accurately predict the net energy value of this type of feed raw materials for growing pigs. Summary of the invention

[0005] The embodiment of the present invention provides a method for predicting the net energy value of feed raw materials on growing pigs. The present invention classifies feed raw materials based on conventional chemical components and can well predict their net energy value on growing pigs, which helps to solve the difficulties of expensive equipment, high professional requirements, large workload, and long time required for determining the net energy value of feed raw materials, and further promotes the realization of accurate formulation of growing pig feed.

[0006] An embodiment of the present invention provides a method for predicting the net energy value of feed raw materials for growing pigs, the method comprising:

[0007] 1) Obtain the conventional chemical composition and net energy value of various feed raw materials;

[0008] 2) clustering the feed raw materials according to their conventional chemical components;

[0009] 3) Perform supervised classification based on the clustering results to obtain the classification standard;

[0010] 4) Classify the feed raw materials according to the classification standard, and use the conventional chemical components of the feed raw materials as input variables, and use the net energy value of the feed raw materials on growing pigs as output variables to establish at least one category of prediction models. In the present invention, after clustering and classifying the conventional chemical components of various feed raw materials, a classification standard is obtained, and then the conventional chemical components of various types of feed raw materials are used as input variables, and the measured net energy value on growing pigs is used as the output variable, to establish a feed raw material net energy prediction model with a wide range of applications, eliminating many complex steps in the traditional model to obtain the net energy value of feed raw materials, greatly saving time, manpower, and economic costs, providing great convenience for the determination of the net energy value of feed raw materials, and providing support for the precise preparation of feed formulas.

[0011] According to the present invention, the multiple feed raw materials are feed raw materials with multiple chemical components (including starch and crude protein). The present invention classifies any feed raw material and inputs its conventional chemical components into the prediction model of the category to directly and accurately infer the net energy value of the feed raw material on growing pigs, which helps to solve the difficulties of expensive equipment, high professional requirements, large workload, long time consumption, etc. required for determining the net energy value of feed raw materials, and further promotes the realization of accurate formulation of growing pig feed.

[0012] According to the method for predicting the net energy value of feed raw materials for growing pigs provided by an embodiment of the present invention, the conventional chemical components include one or more of crude protein, starch, crude fat, neutral detergent fiber, acid detergent fiber, crude ash and total energy. In the present invention, the above-mentioned chemical components used are the most generally accepted raw material composition concepts in feed science, and are simple and easy to measure, with accurate measurement accuracy. Compared with directly measuring the net energy value of feed raw materials, it is simpler, more accurate and faster. Using the above method to use conventional chemical components to accurately predict the net energy value of feed raw materials can greatly lower the measurement threshold and reduce the actual workload, so that the net energy system can be better applied to pig production. The application of this most accurate energy system will also bring good economic benefits to actual production.

[0013] According to the method for predicting the net energy value of feed raw materials for growing pigs provided by an embodiment of the present invention, in step 1), the feed raw materials are added to the basal diet as a test diet. During the test period, at least one pig is raised in a respiration calorimetry chamber, and the net energy value of the measured feed raw materials is obtained by feeding the test diet and collecting samples and data.

[0014] According to the method for predicting the net energy value of feed raw materials for growing pigs provided by an embodiment of the present invention, in step 2), the clustering method is normal mixture clustering, and the number of clusters is 3.

[0015] According to the method for predicting the net energy value of feed raw materials for growing pigs provided by an embodiment of the present invention, in step 3), the supervised classification method is recursive segmentation, and the classification standard is: feed raw materials with starch ≥56.75% are category 1; feed raw materials with starch <56.75% and crude protein >26.35% are category 2; feed raw materials with starch <56.75% and crude protein <26.35% are category 3.

[0016] According to the method for predicting the net energy value of feed raw materials for growing pigs provided by an embodiment of the present invention, in step 4), the established prediction model includes:

[0017] The net energy value of the feed ingredients of category 1 for growing pigs = -31.49 + 1.754 × total energy + 0.149 × crude protein + 0.125 × starch + 0.139 × neutral detergent fiber – 0.305 × acid detergent fiber;

[0018] The net energy value of the feed ingredients of category 2 for growing pigs = -6.815 + 0.986 × total energy - 0.137 × neutral detergent fiber + 0.051 × acid detergent fiber;

[0019] The net energy value of the Category 3 feed ingredients for growing pigs = -22.28 + 1.751 x gross energy - 0.108 x crude protein - 0.133 x crude fat + 0.076 x starch - 0.126 x acid detergent fiber.

[0020] In the present invention, there are many types of feed raw materials, and the huge differences in sources, textures, and chemical components make their digestion processes in pigs different. Therefore, in traditional feed science, feed raw materials are artificially divided into energy feed raw materials, protein feed raw materials, and fiber feed raw materials based on empirical principles. Classifying feed raw materials according to chemical components can better summarize the digestion rules of feed raw materials and more accurately predict their net energy values ​​on growing pigs. However, traditional artificial classification does not have a scientific and accurate classification standard. Therefore, the present invention adopts a clustering method to obtain a classification standard, and uses the prediction model established by the three categories after classification to more accurately predict the net energy values ​​of feed raw materials from different sources and different conventional chemical components on growing pigs, thereby more accurately applying the net energy system.

[0021] The method for predicting the net energy value of feed raw materials for growing pigs provided in an embodiment of the present invention also includes the following steps: obtaining the conventional chemical components of the feed raw materials to be tested; determining the category to which the feed raw materials to be tested belong according to the conventional chemical components and classification standards of the feed raw materials to be tested; inputting the conventional chemical components of the feed raw materials to be tested into the prediction model of the feed raw materials of the category to which they belong, to obtain a net energy prediction value.

[0022] According to the method for predicting the net energy value of feed raw materials for growing pigs provided by the embodiment of the present invention, the R 2 (coefficient of determination), RMSE (root mean square error) and Akaike information criterion (AIC) were used as criteria to judge the accuracy of the prediction model.

[0023] The method for predicting the net energy value of feed raw materials on growing pigs provided in an embodiment of the present invention further includes: verifying the prediction model using external data; preferably, the external verification data is derived from conventional chemical components of other feed raw materials outside the model establishment and the corresponding net energy values ​​on growing pigs.

[0024] According to the method for predicting the net energy value of feed raw materials for growing pigs provided by an embodiment of the present invention, the net energy value of multiple feed raw materials is determined by a growing pig digestion metabolism and energy balance test in a respiratory calorimeter chamber; and / or, the clustering, classification and prediction model is established using the JMP Pro14.0 program.

[0025] According to an embodiment of the present invention, a method for predicting the net energy value of feed raw materials for growing pigs is provided, the method comprising:

[0026] 1) Obtain common chemical components of various feed raw materials;

[0027] 2) obtaining the net energy values ​​of the plurality of feed raw materials;

[0028] 3) clustering the plurality of feed raw materials according to their conventional chemical components;

[0029] 4) Perform supervised classification based on the clustering results to obtain classification standards;

[0030] 5) classifying a plurality of feed raw materials according to the classification standard, and using the conventional chemical components of each type of feed raw material as input variables and its net energy value on growing pigs as output variables, to establish at least one prediction model for each category;

[0031] 6) Obtaining the conventional chemical components of the feed raw materials to be tested;

[0032] 7) Determine the category of the feed raw material to be tested according to its conventional chemical components and classification standards;

[0033] 8) Inputting the conventional chemical components of the feed raw material to be tested into the prediction model of the feed raw material of this category to obtain the predicted value of net energy.

[0034] The embodiment of the present invention also provides a model for predicting the net energy value of feed raw materials on growing pigs, and the prediction model includes:

[0035] The net energy value of Category 1 feed ingredients for growing pigs = -31.49 + 1.754 × gross energy + 0.149 × crude protein + 0.125 × starch + 0.139 × neutral detergent fiber – 0.305 × acid detergent fiber;

[0036] The net energy value of Category 2 feed ingredients for growing pigs = -6.815 + 0.986 × gross energy - 0.137 × neutral detergent fiber + 0.051 × acid detergent fiber;

[0037] The net energy value of Category 3 feed ingredients for growing pigs = -22.28 + 1.751 × gross energy – 0.108 × crude protein – 0.133 × crude fat + 0.076 × starch – 0.126 × acid detergent fiber;

[0038] Category 1 is a feed material with starch ≥56.75%; Category 2 is a feed material with starch <56.75% and crude protein >26.35%; Category 3 is a feed material with starch <56.75% and crude protein <26.35%. In the present invention, the model is used to classify any feed material, and then the net energy value of the feed material on growing pigs is directly and accurately inferred by inputting its conventional chemical components into the prediction model of this category, which helps to realize the rapid prediction of the net energy value of the feed material on growing pigs and provides strong support for the realization of accurate formulation of pig feed.

[0039] The embodiment of the present invention also provides a method for predicting the net energy value of feed raw materials on growing pigs or an application of the model for predicting the net energy value of feed raw materials on growing pigs in the precise formulation of pig feed.

[0040] The beneficial effects of the present invention are at least that: the present invention objectively classifies and establishes a prediction model based on the relationship between the conventional chemical components of feed raw materials and their net energy values ​​for the first time, and provides a method for predicting the net energy value of feed raw materials on growing pigs based on their conventional chemical components. First, the category to which the feed raw materials belong is clarified according to their conventional chemical components and classification standards, and their conventional chemical components are input as input variables into the net energy prediction model of their corresponding categories, so that their net energy prediction values ​​can be directly and accurately obtained, thereby realizing the rapid prediction of the net energy value of feed raw materials on growing pigs, lowering the threshold for determining the net energy value, and reducing the complicated determination process, which provides a huge boost to the application of the net energy system and provides strong support for the realization of precise formulation of pig feed. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 A flow chart of a method for predicting the net energy value of feed raw materials on growing pigs provided by the present invention;

[0043] Figure 2 A schematic diagram of the results of verifying the net energy prediction using external data provided by the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] If no specific technology or conditions are specified in the examples, the technology or conditions described in the literature in this field or the product instructions are used. If the manufacturer is not specified for the instruments used, they are all conventional products that can be purchased through regular channels. The methods are conventional methods unless otherwise specified, and the raw materials can be obtained from public commercial channels unless otherwise specified. If no specific technology or conditions are specified in the examples, the technology or conditions described in the literature in this field or the product instructions are used.

[0046] In the embodiment of the present invention, the index system for measuring the effective energy value of pig feed raw materials mainly includes digestible energy, metabolic energy and net energy system. Compared with the digestible energy and metabolic energy systems, the net energy system can more truly reflect the energy utilization efficiency of the feed, and is the only effective energy system that considers the energy requirements of animals and the energy value of feed on the same basis. Specifically, gross energy refers to the energy released by the complete combustion of organic matter in the feed. Digestible energy is the energy remaining after deducting the undigested energy excreted through feces from the total energy. Metabolic energy is the energy remaining after deducting the energy of urine and combustible gases in the digestive tract from the digestive energy. Net energy is the energy remaining after deducting heat gain from metabolic energy. Heat gain refers to the heat generated by animals during metabolism, feeding, digestion and other physiological processes of the body, which is generally calculated by subtracting net maintenance energy from total heat production. Net maintenance energy cannot be measured directly, and is generally approximated by fasting heat production.

[0047] In the embodiments of the present invention, the content of the conventional chemical components is determined using national standards; the net energy value of different feed raw materials is determined by a growing pig digestion metabolism and energy balance test in a respiratory calorimeter chamber; and the clustering, classification and prediction model is established using the JMP Pro 14.0 program.

[0048] In the embodiment of the present invention, the calculation formulas for gross energy, digestible energy, metabolic energy, fecal energy, urine energy, net energy, heat gain and fasting heat production are as follows:

[0049] Digestible energy = gross energy – fecal energy;

[0050] Metabolic energy = digestive energy – urine energy – methane energy;

[0051] Net energy = metabolizable energy – total heat production + fasting heat production;

[0052] Heat expenditure = total heat production – fasting heat production;

[0053] Heat production = 16.1753×O2(L)+5.0208×CO2(L)–2.1673×CH4(L)–5.9873×urine nitrogen(g).

[0054] Example 1 Method for predicting the net energy value of feed raw materials for growing pigs

[0055] This embodiment provides a method for predicting the net energy value of feed raw materials for growing pigs, the steps are as follows:

[0056] (1) First, obtain the conventional chemical composition of various feeds and their net energy values ​​for growing pigs.

[0057] The conventional chemical components of feed raw materials are obtained by the following methods:

[0058] Crude protein, crude fat, crude ash and neutral detergent fiber were determined according to the national standards of the People's Republic of China GB / T 6432-1994, GB / T 6433-2006, GB / T 6438-2007 and GB / T 20806-2006, respectively. Acid detergent fiber was determined according to the agricultural industry standard of the People's Republic of China NY / T1459-2007. Starch was determined according to AOAC (2007), and gross energy was determined according to the method recommended by international standard ISO9831:1998 using an oxygen bomb calorimeter (Parr6300Calorimeter, Moline, IL).

[0059] The net energy value of feed ingredients for growing pigs is obtained through digestion metabolism and energy balance tests of growing pigs, as follows:

[0060] To ensure the rigor of the test process and data, the animal test was conducted in an animal test center in Hebei. Each feed raw material was added to the basic diet in a specific proportion based on its feed properties and palatability as the test diet. The basic diet was a corn-based diet or a corn-soybean meal-based diet. Each test diet was repeated 6 times, with 1 pig in each repeat. The weight of the test pigs should be similar. The test lasted for 16 days. The first 7 days were the pre-feeding period. The test was conducted in metabolic cages and fed twice a day at 2200kJ ME / kg BW. 0.6 Feed. On the 8th day, transfer the test pigs to the respiratory calorimetry chamber and adapt for 1 day. From the 9th to the 13th day, collect all the excrement and urine and measure the O2 inhalation, CO2 exhalation and CH4 exhalation of the test pigs. 14th to 15th day is the transition period from the feeding period to the fasting period, and 16th day is the fasting period. The O2 inhalation, CO2 exhalation and CH4 exhalation 8 hours before the end of the fasting period are used to calculate the fasting heat production.

[0061] Feces and urine samples were collected using the full feces collection method. Feces and urine were collected once in the morning, noon, and evening every day, and the feces and urine samples collected daily were stored in a -20℃ refrigerator. After the 5-day collection period, all feces samples were mixed evenly, and some feces samples were dried in a 65℃ oven for 72 hours, and crushed after 24 hours of moisture recovery for testing. During the 5-day feces and urine collection period, 10% HCl was used to fix nitrogen in the urine, and 20 mL of 10% HCl was added to every 1000 mL. Finally, all urine was mixed evenly and 10% of the sample was placed in a -20℃ refrigerator for testing.

[0062] The determination of gross energy, fecal energy, and urine energy was carried out according to the method recommended by the international standard ISO9831:1998, using an oxygen bomb calorimeter (Parr 6300 Calorimeter, Moline, IL). O2 was determined using a paramagnetic oxygen analyzer (Oxymat 6E, Siemens, Munich, Germany). CO2 and CH4 were determined using an infrared analyzer (Ultramate 6E, Siemens, Munich, Germany).

[0063] Metabolizable energy of experimental diet (MJ / kg, dry matter basis) = (gross energy ingested, MJ – gross energy excreted, MJ – urine energy, MJ – methane energy, MJ) / dry matter intake, kg;

[0064] Net energy of experimental diet (MJ / kg, dry matter basis) = (metabolizable energy of food, MJ – total heat production, MJ + fasting heat production, MJ) / dry matter intake, kg;

[0065] Heat production (kJ) = 16.1753 × O2 (L) + 5.0208 × CO2 (L) – 2.1673 × CH4 (L) – 5.9873 × urine nitrogen (g);

[0066] After calculating the net energy value of the test diet and the basic diet based on the above contents, the net energy value of a certain feed raw material for growing pigs can be calculated based on the proportion of the feed raw material replacing the basic diet.

[0067] The net energy value of feed ingredients for growing pigs = (net energy value of test diet – (100% – X%) × net energy value of basic diet) / X%, where X% is the percentage of the energy supply component of the basic diet replaced by the feed ingredient in the test diet.

[0068] (2) Cluster and classify various feed raw materials to obtain classification standards.

[0069] A total of 39 feed raw materials were clustered and classified, specifically, including corn, naked oats, brown rice, wheat, sorghum, millet, barley, partially hulled barley, hulled millet, soybean meal, rapeseed meal, cottonseed meal, peanut meal, sunflower meal, corn dry distillers grains with solubles (DDGS), wheat bran, corn germ meal, rice bran meal, rice bran, oat bran, corn husks, palm meal, corn gluten feed, etc.

[0070] The data were statistically analyzed using JMP Pro 14.0. The clustering method was normal mixture, with starch, neutral detergent fiber, and crude protein as clustering criteria, the number of clusters set to 3, the number of processes set to 30, and the maximum number of iterations set to 300. After clustering, there were 3 categories in total. In order to avoid subjective factors affecting the classification criteria, the 3 categories were named category 1, category 2, and category 3; the category was used as the response, and starch, neutral detergent fiber, and crude protein were used as independent variables for supervised classification. The method selected was recursive segmentation, and the final classification criteria were: feed raw materials with starch ≥56.75% were classified as category 1; feed raw materials with starch <56.75% and crude protein >26.35% were classified as category 2; feed raw materials with starch <56.75% and crude protein <26.35% were classified as category 3.

[0071] (3) Establish a prediction model for the net energy value of feed raw materials for growing pigs.

[0072] The data of three categories of feed raw materials were statistically analyzed using the fitting model in JMP Pro 14.0 statistical software. The stepwise regression method in the fitting model program was used to establish a prediction model for the net energy value of feed raw materials, in which the coefficient of determination (R2), root mean square error (RSME) and Akaike information criterion (AIC) were used as indicators to evaluate the most suitable prediction model. The prediction model with the largest R2, smallest RSME and AIC was the most suitable model.

[0073] The optimal net energy prediction model established is as follows:

[0074] Net energy value of Category 1 raw materials for growing pigs = -31.49 + 1.754 × gross energy + 0.149 × crude protein + 0.125 × starch + 0.139 × neutral detergent fiber – 0.305 × acid detergent fiber;

[0075] Net energy value of Category 2 raw materials for growing pigs = -6.815 + 0.986 × total energy - 0.137 × neutral detergent fiber + 0.051 × acid detergent fiber;

[0076] The net energy value of Category 3 raw materials for growing pigs = -22.28 + 1.751 × gross energy – 0.108 × crude protein – 0.133 × crude fat + 0.076 × starch – 0.126 × acid detergent fiber.

[0077] Example 2 Validation of the prediction model

[0078] After the prediction model was established, the net energy prediction model of feed raw materials was used for external verification. The verification data included 3 types of feed raw materials. The category of the raw material to be tested was determined according to the conventional chemical components and classification standards of the feed raw materials to be tested, and its conventional chemical components were used as input variables to input into the prediction model of this category to obtain the predicted value of the net energy value of the feed raw material on growing pigs. Table 1 shows the conventional chemical components and classification categories of the 3 feed raw materials. Table 2 shows the net energy prediction results obtained for the 3 feed raw materials using the above prediction model. The data results are as follows:

[0079] Raw material 1 to be tested: The gross energy of wheat is 18.41MJ / kg, the crude protein is 15.37%, the crude fat is 1.78%, the starch is 67.90%, the neutral detergent fiber is 13.83%, the acid detergent fiber is 3.52%, the crude ash is 1.87%, the classification category is category 1, the measured net energy value is 12.52MJ / kg, the predicted net energy value is 12.43MJ / kg, the absolute error is -0.09MJ / kg, and the relative error is -0.71%.

[0080] Raw material 2 to be tested: Northwest rapeseed meal has a total energy of 19.47MJ / kg, crude protein of 39.92%, crude fat of 2.58%, starch of 3.02%, neutral detergent fiber of 32.27%, acid detergent fiber of 21.50%, crude ash of 8.24%, and classification category of 2. The measured net energy value is 7.98MJ / kg, and the predicted net energy value is 8.37MJ / kg. The absolute error is 0.39MJ / kg, and the relative error is 4.94%.

[0081] Raw material 3 to be tested: The total energy of rice bran is 21.78MJ / kg, the crude protein is 15.82%, the crude fat is 2.58%, the starch is 29.06%, the neutral detergent fiber is 22.06%, the acid detergent fiber is 9.45%, the crude ash is 8.80%, the classification category is category 3, the measured net energy value is 12.02MJ / kg, the predicted net energy value is 12.11MJ / kg, the absolute error is 0.09MJ / kg, and the relative error is 0.78%.

[0082] The absolute errors of the prediction results of all samples were within 0.5 MJ / kg, the relative errors were within 5%, and the chemical components of the three feed raw materials were very different, which fully demonstrated the high accuracy and wide applicability of the net energy prediction model proposed in the embodiment of the present invention. Figure 2 The following is a graph showing the net energy prediction results of three feed ingredients on growing pigs. The dotted line in the figure is the actual value of net energy. It can be seen that the net energy prediction values ​​of the three feed ingredients on growing pigs are all near the straight line, which also shows that the net energy prediction model has high accuracy and wide applicability.

[0083] Table 1 Common chemical components and classification categories of feed raw materials

[0084]

[0085] Table 2 Net energy prediction results of feed raw materials using the above prediction model

[0086]

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the net energy value of feed ingredients for growing pigs, include: 1) Obtain the conventional chemical composition and net energy value of various feed raw materials; 2) clustering the feed raw materials according to their conventional chemical components; 3) Perform supervised classification based on the clustering results to obtain the classification criteria; 4) classifying the feed raw materials according to the classification standard, taking the conventional chemical components of the feed raw materials as input variables, and taking the net energy value of the feed raw materials on growing pigs as output variables, to establish a prediction model for at least one category; In step 3), the supervised classification method is recursive segmentation, and the classification standard is: feed raw materials with starch ≥56.75% are classified as category 1; feed raw materials with starch <56.75% and crude protein >26.35% are classified as category 2; feed raw materials with starch <56.75% and crude protein <26.35% are classified as category 3; In step 4), the prediction model established includes: Net energy value of the feed ingredients of category 1 for growing pigs = -31.49 + 1.754 × gross energy + 0.149 × crude protein + 0.125 × starch + 0.139 × neutral detergent fiber – 0.305 × acid detergent fiber; The net energy value of the Category 2 feed ingredients for growing pigs = -6.815 + 0.986 × gross energy - 0.137 × neutral detergent fiber + 0.051 × acid detergent fiber; The net energy value of the Category 3 feed ingredients for growing pigs = -22.28 + 1.751 × gross energy – 0.108 × crude protein – 0.133 × crude fat + 0.076 × starch – 0.126 × acid detergent fiber.

2. The method for predicting the net energy value of feed raw materials for growing pigs according to claim 1, It is characterized in that The conventional chemical components include one or more of crude protein, starch, crude fat, neutral detergent fiber, acid detergent fiber, crude ash and gross energy.

3. The method for predicting the net energy value of feed raw materials for growing pigs according to claim 1, It is characterized in that In step 1), feed raw materials are added to the basic diet as the test diet. During the test period, at least one pig is kept in a respiration calorimeter chamber, and the net energy value of the feed raw materials is obtained by feeding the test diet and collecting samples and data.

4. The method for predicting the net energy value of feed raw materials for growing pigs according to claim 1, It is characterized in that In step 2), the clustering method is normal mixture clustering, and the number of clusters is 3.

5. The method for predicting the net energy value of feed raw materials for growing pigs according to claim 1, It is characterized in that The determination coefficient R of the prediction model is 2 , root mean square error RMSE and Akaike information criterion AIC are used as criteria for judging the accuracy of the prediction model.

6. A method for predicting the net energy value of feed raw materials for growing pigs according to any one of claims 1 to 5, It is characterized in that Also includes: Obtain the conventional chemical composition of the feed raw materials to be tested; Determine the category of the feed raw material to be tested according to its conventional chemical components and classification standards; The conventional chemical components of the feed raw material to be tested are input into the prediction model of the feed raw material of the corresponding category to obtain the net energy prediction value.

7. A method for predicting the net energy value of feed raw materials for growing pigs according to any one of claims 1 to 5, It is characterized in that Also includes: The prediction model is validated using external data.

8. The method for predicting the net energy value of feed raw materials for growing pigs according to claim 7, It is characterized in that The external validation data were derived from conventional chemical compositions of other feed ingredients outside the model establishment and their corresponding net energy values ​​for growing pigs.

9. Application of the method for predicting the net energy value of feed raw materials for growing pigs according to any one of claims 1 to 8 in the precise formulation of pig feed.

Images